We use a hierarchical Bayesian logit model to assess the probability of training during the sixth wave, with a lagged dependent variable as a predictor. Results suggest that training participation is path-dependent and access to training is limited for people who have not trained previously. We find a relationship between a macrostructural context and path dependency during training. An interaction between macro-predictors and the lagged dependent variable shows that access to training is greater in countries with stronger knowledge economies, stronger emphasis on education, and a proactive ageing climate. The size of the welfare system plays no role. These results have implications for policies that address problems of cohesion, active ageing and, adult learning. We argue that improvement of accessibility requires adequate measures that take into account path dependency and life course perspective.